OFB Community Mailing Lists

[Eduardo Sánchez]

On Tuesday 05 May 2009 10.05.58 Karl Kleinpaste wrote:
> Eduardo Sánchez <lists at sombragris.org> writes:
> > No I won't. There's just one reminder: gcc-2.96 in RH 7.0.
>
> Let's see...  I could have sworn I addressed this, didn't I?  Hm. 
> Let
>
> me look -- yup, there it is:
> >> There have been
> >> good reasons to complain at them now and again
>
> And while you're entitled to believe that gcc 2.96 was a reason for
> such complaint, I would argue that you're wrong even on technical
> grounds, much less social grounds.
>
> More to the point, however, RH7 was a decade in the past.

Look, I just mentioned what made Red Hat famous. They might be doing 
really fine now. But they got quite a reputation for those practices.

And as for the technical grounds of the unreleased and producer-of-
broken-code gcc-2.96, let's leave it at that. I don't want to argue the 
squareness of a circle.

I also don't want to argue about a really annoying bug in Qt 2.x that 
broke deadkeys. In that way, no RH user could ever write in Western 
European languages (iso 8859-1 languages!) in KDE. The responsible was a 
patch applied by RH, and RH only. Everybody else had a perfectly working 
Qt/KDE, but RH couldn't. At the bug report, all they could manage to say 
is, "use the Compose key".

> It appears
> you don't believe that a sinner could ever convert.  Got any example
> that hasn't already ossified, and grown a fine sheen of moss besides?
> Since you said "there's just one" above, evidently not.
>
> Such whining is still over the top, and not reflective of what I see
> in the trenches, because that's where I am today.  Most of Fedora
> package management is quite good.  Yes, they do patches beyond
> upstream content. Funny, they have good reason for those patches just
> about all the time. It's not like they wake up in the morning and
> decide, "Hey, let's screw over package XYZ today."  Yes, they test. 
> Yes, they sometimes let a bug slip through (e.g. about a month ago,
> NetworkManager's SIGSEGV death in a conflict with VMware).  This is
> not an indication of systemic flaw; this is an indication of an
> occasional -- very occasional -- QA glitch.
>

Well, let's hope they improved. In fact, I don't hear a lots of 
complaints related to Red Hat/Fedora nowadays so they might be in fact 
much better than they were before. But what I claimed as the "RH/Fedora 
philosophy" was what made RH/Fedora (in)famous, and it's, IMHO, what is 
creeping now into Ubuntu. 

Let's stop this. I don't think this is a point worth arguing about, and 
I understand your efforts toward improving RH/Fedora quality. Everyone 
will figure out what truth is in our statements.

Blessings,

Eduardo

-- 
Eduardo Sanchez, B. Th.
Traductor Público Inglés-Español
http://shadow.sombragris.org
--------------------------------------------------------------
 Would but some winged Angel ere too late
 Arrest the yet unfolded Roll of Fate,
   And make the stern Recorder otherwise
 Enregister, or quite obliterate!

	-- The Rubaiyat of Omar Khayyam
	   

--------------------------------------------------------------